Download Participating convertible preferred stock in venture capital

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Document related concepts

Mergers and acquisitions wikipedia, lookup

Shareholder value wikipedia, lookup

Special-purpose acquisition company wikipedia, lookup

Startup company wikipedia, lookup

Private equity secondary market wikipedia, lookup

Private equity wikipedia, lookup

Stock exchange wikipedia, lookup

Stock wikipedia, lookup

History of private equity and venture capital wikipedia, lookup

Private equity in the 2000s wikipedia, lookup

Venture capital wikipedia, lookup

Initial public offering of Facebook wikipedia, lookup

Initial public offering wikipedia, lookup

Transcript
JBV-05675; No of Pages 16
Journal of Business Venturing xxx (2013) xxx–xxx
Contents lists available at SciVerse ScienceDirect
Journal of Business Venturing
Participating convertible preferred stock in venture capital exits☆
Sridhar Arcot ⁎
ESSEC Business School, Department of Finance, Avenue Bernard Hirsch, B.P. 50105, 95021 Cergy-Pontoise Cedex, France
a r t i c l e
i n f o
Article history:
Received 13 July 2011
Received in revised form 5 June 2013
Accepted 5 June 2013
Available online xxxx
Field Editor: G. Cassar
JEL classification:
G24
G32
a b s t r a c t
This paper develops a theory of the participating convertible preferred (PCP) stock commonly
used in venture capital settings. I show that the participation and convertibility features of PCP
stock can be used to reduce information asymmetry between the venture and potential
investors at the time of exit. Further, the convertibility feature of PCP helps in alleviating the
problem of insufficient entrepreneurial effort. I then derive implications for the two most
common types of exits in venture capital—initial public offerings and trade sales—and explain
how US venture capital markets differ from other VC markets.
© 2013 Elsevier Inc. All rights reserved.
Keywords:
Participating convertibles
Exits
Signaling
IPOs
TS
1. Executive summary
Venture capital is an important source of equity financing for innovative ventures. Venture capitalists typically use convertible
preferred equity to finance such ventures. Investors of convertible preferred equity have the option of either holding a debt-like
claim-preferred equity, or converting into common equity. The literature has analyzed the main trade-offs affecting the two alternatives.
However, convertible preferred equity can come with special features. With a few exceptions, scholars have not delved into the specific
details of these securities. This paper examines one such convertible preferred equity with distinct features—participating convertible
preferred (PCP) stock.
Participating convertible preferred stock gives its holders preference in dividend payments and at the same time allows them to
participate in excess earnings (i.e. the cash flows to which equity is entitled to after all debt and preferred claims have been satisfied)
along with the common stockholder. PCP holders thus concurrently hold both a debt-like claim (preferred equity) as well as an equity
claim (participation rights). However, PCP holders lose their preferred rights if they convert into common stock. The specific question that
I address in this paper is why are venture capitalists willing to convert their PCP stock into common equity and give up their preferred
rights?
☆ I am grateful to my supervisor the late Antoine Faure-Grimaud for the helpful comments and discussions during every step of this paper. I would also like to
thank David Webb for his encouragement. I have received useful comments from Stefan Ambec, Carsten Bienz, Marco Da Rin and Uwe Walz as well as from the
participants at the Second RICAFE Conference (held at Frankfurt in October 2004) and the Portuguese Finance Network (held in Azores during 2010). For the
helpful suggestions I thank Margaret Bray and the participants at the various seminars of the London School of Economics and Felix Papier from ESSEC. Finally, I
thank the three anonymous referees and Gavin Cassar, the field Editor for their insightful comments. Financial support from the FMG, European Union grant no
HPSE-CT-2002-00140 and CERESSEC is gratefully acknowledged. All remaining errors are my own.
⁎ Tel.: +33 1 34 43 30 77.
E-mail address: [email protected]
0883-9026/$ – see front matter © 2013 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.jbusvent.2013.06.001
Please cite this article as: Arcot, S., Participating convertible preferred stock in venture capital exits, Journal of Business Venturing
(2013), http://dx.doi.org/10.1016/j.jbusvent.2013.06.001
2
S. Arcot / Journal of Business Venturing xxx (2013) xxx–xxx
To answer this question, I propose a signaling model for PCP stock based on its role in venture capital exits. The two major
forms of exits observed in venture capital are the initial public offerings (IPOs) and the trade sale (TS). Typically, a PCP stake is
converted into common equity during an IPO exit but is not converted in a TS exit. My model shows that VCs can signal the quality
of their venture in an IPO, by converting their PCP stake into common equity and giving up some of their cash flow rights.
Signaling is of particular importance in an IPO because new investors are relatively uninformed about the venture. In contrast,
potential trade buyers are more likely to be well-informed since they are usually industry peers.
When exit is through an IPO, the entrepreneur retains control of the firm. Thus, when the firm value is high, an IPO exit
rewards the entrepreneur and should be the preferred exit route. However, the VC may be reluctant to take that route given that
investors in an IPO are less informed and the VC may not get the full value for his stake. So it is precisely when the firm value is
high that the VC may prefer to target investors who are more informed and thus less costly—in other words to exit through a trade
sale. However, the interests of VCs and entrepreneurs are more easily aligned when the latter convert their PCP stakes into
common shares and exit through an IPO.
With the help of the model, I confirm some well-known empirical observations. The model first predicts that, the greater a market's
informational efficiency, the greater the possibility of signaling. There is empirical evidence to suggest that the US and UK markets are
informationally more efficient compared to all other markets. Hence we are more likely to observe exits through IPOs in US and UK than
in other markets, which is borne out by the evidence. Secondly, the higher the value of reputational benefits that a VC derives by exiting a
venture through an IPO, the higher is the probability of such an exit. This has been empirically confirmed by Gompers (1996).
Venture capitalists investing in start-ups use sophisticated financial instruments to structure their investments. This paper
provides a rationale, for the use of one such financial instrument—PCP stock, based on the venture capitalist's exit decision. In doing
so, it makes a connection between the exit route and entrepreneurial effort. The paper thus highlights factors that have direct
implications for the incentives of venture capitalists to invest in ventures and entrepreneurs to exert effort to make them a success.
2. Introduction
Venture capitalists' (VCs) investment in a new venture typically takes the form of convertible preferred stock. Investors in
convertible preferred (CP) either hold debt-like preferred equity or have the option of converting into common equity. The literature
has documented the extensive use of convertible securities in venture capital contracting (Kaplan and Stromberg (, 2003)). While
analyzing the use of CP stock the theoretical work has focused mostly on the plain-vanilla form of these securities and concludes that
it is an optimal incentive structure between the entrepreneur and the VC (Da Rin et al. (2011)). In practice, convertibles do not always
have this simple structure and come in many different flavors (Metrick and Yasuda (2011)). This paper analyzes one such
variant, observed frequently in venture capital contracting, participating convertible preferred (PCP) stock which is a CP stock with
participation rights.
Participation rights entitle the VC, in the event of sale or liquidation, to a liquidation preference plus a pro rata share of what remains
to be paid to common shareholders. Thus, upon sale or liquidation, participating preferred shareholders have a debt-like claim equal to
their liquidation preference plus a common shareholder's claim. In contrast, holders of nonparticipating convertible preferred shares
either receive the liquidation preference payable on the preferred stock or they convert their shares to common stock and share pro rata
with common shareholders. I give a simple numerical example to illustrate these features. Assume that a VC's investment entitles him
to $5 million from a given venture in the form of a CP stock, that is convertible into 50% of the common equity. Now suppose the
company is liquidated for $12 m. The VC can then either (a) convert his stake to common equity and receive 50% of the proceeds (i.e.
$6 m) or (b) not convert and receive his preferred proceeds (i.e. $5 m). Let us further assume now that the VC holds a PCP with
participation rights on an as-converted basis (50% in this case). If the VC converts his PCP stake to common equity then he is entitled to
$6 m (i.e. 50% of $12 m), but if he chooses not to convert then he is entitled to receive $5 m (his preferred claim) plus shares to the
extent of 50% in the remaining equity pool of $7 m ($12 m minus $5 m), thus giving him a total of $8.5 m (i.e. $3.5 m plus $5 m) rather
than $6 m from converting CP stock. Thus the stockholders cash flow rights vary depending on whether or not the stake is converted. In
this example, the VC's payoff is higher in case of nonconversion (conversion) in the presence (absence) of the participation feature.
PCP stock is routinely used in venture capital contracts. Kaplan and Strömberg (2003) report that nearly 80% of all venture
contracts use convertible preferred stock and that in nearly half of those cases the stock is participating. In venture capital
contracts, PCPs are structured in such a way that the allocation of cash flow rights varies depending on the type of exit. The two
most common types of exit observed in venture capital are an initial public offering (IPO) or a trade sale (TS)—in which the
company is sold either to a trade buyer or acquired by another company. Most venture capital investment agreements explicitly
treat TS as a liquidation event, in which case the VCs retain both their participation and preferred rights whereas the same
agreements usually stipulate automatic conversion of the convertible stake into common equity if exit is via an IPO. By giving up
their preferred rights during an IPO, the VCs are forsaking a substantial portion of their cash flows. The question that I address in
this paper is: why are the VCs willing to give up their preferred rights in the case of an IPO but not in the case of a TS?
I propose that VCs use PCP stock conversion as a signal of the firm's quality. My assumption is that signaling is especially important
in a public offering since the new shareholders are relatively uninformed about the venture's value. In contrast, the firms that are
bidding in a TS have the opportunity to conduct due diligence and also tend to be peers from the same industry; this gives them
in-depth knowledge, which makes them relatively well informed. I argue that the VCs convert their stake into common equity—and
accept a lower stake when exit is through a IPO—to signal the quality of the venture to investors. However, for a TS exit, such a costly
signal is not required because target buyers are relatively well informed about the venture's value. Hence the relative costs (versus the
TS scenario) of exiting through an IPO create the possibility of a signaling equilibrium, which good firms can use by converting their
Please cite this article as: Arcot, S., Participating convertible preferred stock in venture capital exits, Journal of Business Venturing
(2013), http://dx.doi.org/10.1016/j.jbusvent.2013.06.001
S. Arcot / Journal of Business Venturing xxx (2013) xxx–xxx
3
PCP stake.1 Such an ex-post equilibrium can also give the entrepreneur ex ante incentives given that she typically is in control of the
venture after a IPO. In contrast, the entrepreneur loses control of her venture after a trade sale. Therefore, provided it occurs when firm
value is high, exit via IPO has the desirable property of rewarding the good entrepreneur with continued control of the venture.
My model works as follows. In the presence of PCP, the VC investing in a good firm would prefer not to convert. Although the IPO
exit route suffers from greater informational asymmetry than the TS route, the latter does have the advantage of allowing VC not to
convert. This means that a IPO is costly to the VC (because of conversion) which provides a role for PCP as a signaling mechanism.
The role of VCs in certifying IPOs has been well documented. Megginson and Weiss (1991) provide empirical evidence for the
certification role of venture capitalists in bringing new issues to market. They show that the support by VCs of the offering firm
certifies the issue's quality through their investment of financial and reputational capital. By comparing the costs of going public
(including underpricing, underwriting spreads, etc.) for a group of VC-backed IPOs with a control sample of non-VC-backed offers,
these authors find that the costs of going public are significantly lower for the VC-backed IPOs. Megginson and Weiss also contend
that the mere presence of the VC is enough to certify the venture. I take this a step further and argue that VCs use their stake
conversion as a further signal of the venture's quality.
Black and Gilson (1998) endorse the view that relinquishing control to entrepreneurs is a means of rewarding good entrepreneurs.
These authors argue that in a US style stock market-based system (unlike a German, bank-based system)—there is an implicit contract
between the VC and the entrepreneur by which, if the venture does well the VC cedes control back to the entrepreneur by exiting
through an IPO. In the case of a TS control of the venture is transferred to the acquirer. Black and Gilson argue that the opportunity to
acquire control incentivizes the entrepreneur much beyond the purely financial gain arising out of an appreciation of her stake. My
model formalizes this argument and in addition confirms the signaling role of PCP stock in the exit decisions of venture capitalists.
To the best of my knowledge, Hellmann (2006) is the only paper that deals explicitly with the use of convertible preferred stock
(and its variants) in venture capital contracts. Hellmann, shows that pure equity is the optimal security for resolving the double moral
hazard problem. Convertible preferred equity preserves balanced incentives if the venture remains independent (after an IPO) yet
allows the VC to extract additional rents if it is acquired (in a TS), thus helping to satisfy his financing constraint. Hellmann's paper thus
focuses on the early-stage investor whereas I focus on the late-stage investor: the strategic buyer or IPO investor. My model therefore
analyzes the “buy” side of the venture market in the presence of asymmetric information, so it complements Hellmann's work by
suggesting an additional reason for the use of these securities—namely, their role in signaling the venture's quality at the time of VC exit.
This paper also contributes to the literature on venture capital exits. Berglöf (1994) and Bascha and Walz (2001) model the
trade-offs between IPO and TS exit routes. In both these papers and also in the paper by Hellmann (2006) there are conflicts
between the entrepreneur and VC on the most appropriate method of exit from a venture. Convertible securities can help the
parties select an optimal exit strategy by suitably allocating control rights. In the model I abstract from this conflict and instead
address the question of how convertibles help resolve information asymmetry and provide ex ante incentives to the
entrepreneur. Other research on VC exits includes the papers by Bienz and Leite (2008), who focus on control issues related to
exit and specifically on the post-IPO monitoring of the entrepreneur and Tykvova (2003), who analyzes the optimal timing of exit.
This paper is also related to the work of Faure-Grimaud and Gromb (2004) and Aghion et al. (2000)) who explore how
liquidity shocks affect a VC's desire to exit an investment. Faure-Grimaud and Gromb (2004) emphasize the effect of stock price
informativeness on exit decisions and incentives. Some of my comparative statics deal with the stock market's ability to price IPOs
accurately and in this sense complements their study.
The rest of the paper proceeds as follows. In Section 3, I describe the model set-up. Section 4 sets out the entrepreneur's first-best
incentives and the venture capitalist's choices. In Section 5, I analyze the various signaling equilibria. Section 6 discusses alternative
explanations and suggests avenues for further research. Section 7 examines the empirical implications and Section 8 concludes.
3. The model
Consider a model with three dates, universal risk neutrality and no discounting. An entrepreneur (“she”) has a project that
needs an investment K. Lacking, the financial means to put up this amount she approaches a competitive venture capitalist (“he”)
to fund this project.
Contracts: The VC receives a proportion f of the cash flows. For now, view the security held by the VC as convertible preferred
equity. This stake also comes with participation rights, and the VC can convert his stake at a cost (as described previously) into a
fraction q (bf) of common equity.2 The entrepreneur is the residual claimant.
The proportion f of PCP comprises two components, the preferred dividend D, and a share q of the remaining cash flows. If the
value of the venture at exit is V then the total payoff to PCP holders without conversion is
f ¼
ðD þ ðV−DÞqÞ ð1−qÞD
¼
þq
V
V
ð1Þ
Upon conversion the PCP stockholder receives qV.
1
See Section 3 for the distinction between “good” and “mediocre” ventures.
The issue of security design is much discussed in the theoretical VC literature. Most conclude (albeit for different reasons) that convertible equity is the
optimal security for VC contracting. This paper approached the issue from a different perspective: we take the equity claim as given and focus our analysis on the
participating and convertible features associated with it.
2
Please cite this article as: Arcot, S., Participating convertible preferred stock in venture capital exits, Journal of Business Venturing
(2013), http://dx.doi.org/10.1016/j.jbusvent.2013.06.001
4
S. Arcot / Journal of Business Venturing xxx (2013) xxx–xxx
For all values of D N 0 and 0 b q ≤ 1 we obtain f N q, which justifies our assumption. This analysis suggests that the proportion
of cash flows f is not fixed and instead will depend on the venture's value V at the time of exit. However, in the analysis I will
maintain the assumption of constant proportions for ease of exposition.3
3.1. Project
After the investment is made at t = 1 the venture proves to be either a good firm with a value of VH with probability p or
mediocre firm with value VL (naturally VH N VL). The entrepreneur provides effort e that affects the venture's likelihood of success.
If she exerts effort then the probability of the venture being good is p (and is 0 otherwise). The effort provided is both
unobservable and costly. The value realization (VH or VL) is observed by the VC and entrepreneur but not by outsiders.
After the venture's value is realized, an investment of I is needed to move the venture forward. This investment is valuable, and its
constant rate of return is x. Thus the firm's overall value after investment is either VH(1 + x) or VL(1 + x) as applies. I assume that
absent investment I, the venture is of no value. Both the good project and the mediocre one exhibit a positive net present value (i.e.
VL(1 + x) ≥ K + I), so by this criterion they are worthy of investment. However, neither the entrepreneur nor the VC can invest in
the project at this stage. The reasoning is as follows. The entrepreneur cannot invest because she is wealth-constrained. The VC does
not invest because he is an early-stage investor. We can think of K and I as (respectively) start-up and late-stage investments
respectively. The start-up investment could actually be a series of smaller amounts k1, k2....., kn which is typically provided in stages
upon the entrepreneur's achievement of certain milestones. In most cases investors who provide start-up financing are distinct from
those who invest at later stages. Note also that VCs may have liquidity needs that force them to exit the investment and leave them
with no funds to inject. Venture capitalists tend to invest in firms through a limited partnership. Such arrangements have a finite life
(typically 10–12 years) before they are dissolved. For this reason VCs thus tend to invest in the first 5–6 years of a partnership;
thereafter, they seek to exit their investments so that the partnership can be dissolved.
3.2. Exits
There are only two methods of raising I: through a trade sale or an initial public offering. In the case of a trade sale (TS), buyers
know V with probability r = 1 whereas in an initial public offering (IPO), investors know V with probability r b 1.
We shall assume that buyers in a TS are more informed (and to simplify the exposition, analyze the case where they are fully
informed) than the buyers in an IPO. From an informational point of view it is well known that shares in an IPO are normally sold
to buyers through an intermediary. Investors in IPOs usually rely on the prospectus, which the intermediary also certifies. In
contrast, trade buyers who make an acquisition offer are given full access to the company's books and management, so they can
conduct due diligence; and in most cases, potential acquirers are from the same industry as the target venture and have extensive
knowledge of that industry. Poulsen and Stegemoller (2008) provide empirical evidence that firms with higher asymmetric
information exit through trade sales (as compared to public offerings) because other companies operating in the same industry
are in a better position to value them. Other papers in the literature make similar assumptions, see for e.g. Chemmanur and
Fulghieri (1999), Cumming and MacIntosh (2003), Bayar and Chemmanur (2011) among others.
Finally, our assumption that exit occurs only via IPO or TS implies assuming that the venture is successful and that its value
exceeds the PCP stock's liquidation value (i.e. V N D).
3.3. Control
The model presented here assumes that the VC is in control and makes the exit decision consistent with the empirical evidence
in Bienz and Walz (2007) who analyze 464 German VC–entrepreneur contracts and find that, over time the VC acquires more exit
rights while relinquishing operational rights.
3.4. Preferences
Both the VC and the entrepreneur may derive private benefits from the project.
The entrepreneur derives a private benefit if she is in control of the project. This is usually the case when the VC exits after an
IPO. I therefore assume that the entrepreneur receives private benefit of control B if the VC exits through an IPO (and receives zero
otherwise). That private benefits of control exist is a well-documented empirical fact, one widely used in the control literature
(see Zingales, 1995). Empirical evidence suggests that entrepreneurs continue to be involved with the firm even after an IPO.
Kaplan, Sensoy and Strömberg (2009) find that at the time of the IPO, 92% of the firms continue to have the founder either as a top
executive or director. Even, practitioners acknowledge the existence of private benefits of control. According to a note by the law
firm Baker and McKenzie on venture capital exit routes4; “An exit through the stock market seems to be favored by management,
since it allows them to remain in place and in control.”
3
In maintaining this assumption I sacrifice some features of the security for the sake of simplicity. However, this modeling choice has no effect on the main
trade-off and hence none on the equilibrium.
4
www.bakernet.com/BakerNet/Practice/Corporate
Please cite this article as: Arcot, S., Participating convertible preferred stock in venture capital exits, Journal of Business Venturing
(2013), http://dx.doi.org/10.1016/j.jbusvent.2013.06.001
S. Arcot / Journal of Business Venturing xxx (2013) xxx–xxx
5
I similarly assume that the VC derives a private benefit if the firm is good and the exit is through an IPO. We can conceive
of this private benefit as a reputation effect. Amit et al. (1998) show that VCs might try to acquire a reputation for
sponsoring IPOs only for high quality ventures. It is therefore reasonable to assume that IPOs are associated with a greater
reputation effect for VCs than trade sales. I therefore assume that the VC's private benefits from a successful IPO are higher
than those from a trade sale. The VC receives a private benefit Z if the venture is good and exit is via an IPO (and zero
otherwise).
3.5. Timeline
The model's timeline is summarized as follows:
It is usual for VC to exit from a venture in multiple stages. A small fraction of VC holdings is sold during the IPO and the bulk of
those holdings is sold sometime thereafter. Venture capitalists typically retain a substantial portion of their equity holdings after
an IPO. Megginson and Weiss (1991) report that, on average, venture capitalists own 36.6% (resp. 26.3%) of the firm prior to (resp.
immediately after) the IPO. The timeline aims to reflect these observations.
4. First-best incentives
4.1. Entrepreneur incentives
Assume that the entrepreneur gets a portion of the cash flows, CH, if the venture is good and valued at VH and CL if the venture
is mediocre and is valued at VL. As discussed above besides share of the cash flows the entrepreneur also receives private benefits
if exit is through an IPO. Then the entrepreneur's incentives to create a venture and exert effort is given by
h
i
h
i
h
i
H
L
L
pμ C þ B þ ð1−pÞγ C þ B ≥e þ γ C þ B
ð2Þ
here p is the probability that the venture is good if the entrepreneur exerts effort, μ (resp. γ) is the probability of an IPO given that
the venture is good (resp. mediocre), B is the entrepreneur's private benefit when the exit is through an IPO and e is the disutility
of effort incurred by the entrepreneur.
The left-hand side of Eq. (2) states that, if the entrepreneur exerts effort, then with probability p the venture is good and with
probability μ the exit is through an IPO—in which case the entrepreneur derives the associated cash flow CH and private benefit B.
Of course even when she exerts effort there is still the probability (1 − p) that the venture does not succeed; in that case the VC
exits via IPO (and the entrepreneur enjoys private benefit B) only with probability γ. The entrepreneur will exert effort if she
expects that doing so will yield greater utility than the cost of the effort e and the cash flow and private benefit she gets from not
exerting effort. If the entrepreneur does not exert effort then the venture will be mediocre and with probability γ she then derives
the cash flow CL and private benefit B, this is the right-hand side of Eq. (2).
We can therefore simplify the entrepreneur's incentive compatibility (IC) condition to
h i
H
L
p μ C þ B −γ C þ B ≥e
ð3Þ
The foregoing clearly indicates that an entrepreneur's incentives to exert effort increases (decreases) with μ (γ), the
probability of an IPO when the venture's performance is good (mediocre), respectively.
Please cite this article as: Arcot, S., Participating convertible preferred stock in venture capital exits, Journal of Business Venturing
(2013), http://dx.doi.org/10.1016/j.jbusvent.2013.06.001
6
S. Arcot / Journal of Business Venturing xxx (2013) xxx–xxx
4.2. VC's choices
Like the entrepreneur the VC also values both cash flows and private benefits. The VC chooses his method of exit. Although IPOs
maximize the entrepreneur's incentives, they are informationally disadvantaged in comparison with a trade sale. Once consequence
of the informational asymmetry between the insiders and the investors is that the VC of a good venture might not get a fair price.
If there were no information asymmetries between the VC and the investors, then the venture would need to offer the
following stakes in return for the investment I. If the venture's value were VH, then the investor would demand a share
I
V H ð1 þ xÞ
if the venture's value is VL then the investor's stake in the venture would be
SH ¼
SL ¼
I
V L ð1 þ xÞ
ð4Þ
ð5Þ
Assume instead that the outside investors hold some prior beliefs α that the quality of the venture is good. Given these beliefs,
an investor's stake would have to be:
I
SA ¼ H
αV þ ð1−α ÞV L ð1 þ xÞ
ð6Þ
Now consider the situation from the perspective of a VC who holds a stake of f, that may be converted prior to raising new
capital into q. If the venture is successful (and so has value VH), then the VC has the following choices. The VC can sell his stake as it
is to investors with the most optimistic beliefs (i.e. the highest α). Because we consider the strategic choice faced by a VC who
knows VH, this amounts to choosing the form of exit involving the best-informed buyers—namely, a trade sale. Another option
available to the VC is to use the exit mode as a means of signaling additional information to the market.
When V = VH it seems appropriate for the VC to exit via IPO because doing so efficiently rewards the entrepreneur. However,
the VC may be reluctant to exit that way if IPO investors are less informed. So it is therefore precisely in cases of high firm value
that the VC will prefer to target a more informed audience—and thus prefer to exit via a trade sale.
Yet it may be possible to reconcile the parties' preferences if the VC can somehow signal the firm's type to potential investors.
In the next Section I explore the various possibilities of signaling.
5. Conversion as a signal
I propose that one way for the VC to signal the market is by converting his PCP stake. To be effective in conveying additional
information to the market the signaling action should exhibit two features; first the signal must be costly; second it must be more
costly for the mediocre type than for the good type. When converting PCP to common stock the VC's resulting stake q is less than
his original stake f. Clearly, then conversion qualifies as a signal in terms of the first criterion. Next we examine the conditions
under which the second criterion is also satisfied.
5.1. Conversion without exit choice
I begin by investigating whether a signaling equilibrium exists if there is only one method of exit available. In that case, VCs
cannot choose their exit route but can still convert their PCP stake as a signal to the market. Denote by rB ∈ {r,1} the probability
that a given investor discovers the true value of V after investing.
The VC may seek to signal his type by converting his preferred stake f to common equity q. I look for a separating equilibrium
in which the VC of a good firm converts whereas the VC of a mediocre one does not. The analysis of this equilibrium is
constructive that is a set of investor beliefs is specified and then a program is constructed that assures that the firms behave
accordingly. I begin by assuming that investors believe that the VC of a good firm converts his stake to common equity in order to
signal the type. Therefore the investors believe the venture is a good if they observe conversion (and otherwise believe it is
mediocre). Based on these beliefs, investors demand a suitable stake in return for their investment. Denote by Zr ∈ {0,Z} the VC's
private benefits, where Zr = Z if the exit is through an IPO.
In order for such an equilibrium to exist, the following incentive compatibility conditions must be satisfied.
h
i
H
H
L
qð1−SH ÞV ð1 þ xÞ þ Z r ≥f ð1−SL Þð1 þ xÞ r B V þ ð1−rB ÞV þ Z r
ð7Þ
h
i
L
L
H
f ð1−SL ÞV ð1 þ xÞ≥qð1−SH Þð1 þ xÞ r B V þ ð1−r B ÞV
ð8Þ
If the venture is good the investors accept a stake SH (Eq. (4)) in return for their investment. Condition (7) simply states that
the VC's utility from the good venture after conversion of his stake into common equity is greater than not converting and being
Please cite this article as: Arcot, S., Participating convertible preferred stock in venture capital exits, Journal of Business Venturing
(2013), http://dx.doi.org/10.1016/j.jbusvent.2013.06.001
S. Arcot / Journal of Business Venturing xxx (2013) xxx–xxx
7
unsure of the price that will be offered by the investors. Investors offer a fair price with probability rB. In line with our assumption,
the VC of the good venture derives private benefit Zr in the event of a successful exit.
Similarly, the condition (8) states that there is no advantage to a mediocre firm's VC mimicking a good firm. By converting his
stake (to mimic the good-venture behavior), the VC of a mediocre venture can secure investment I with a “good venture”
purchase price. After buying a stake in this firm the investor discovers its true value with probability rB. Together, then conditions
(7 and 8) lead to the following statement:
Proposition 1. If the VCs of good and mediocre ventures have only one available exit strategy, then there is no separating equilibrium
in which the good venture's VC can signal its type by converting his PCP stake into common equity.
Proof. See the Appendix A.
Therefore, a VC who cannot choose his exit mode will be unable to signal the venture's type through preferred stock conversion. The
reason is that even though the good venture's VC is prepared to bear (via conversion) the cost of signaling, the mediocre venture's VC can
easily convert and mimic that behavior and thereby render such signals less informative. Converting costs a VC to lose (f-q)% of his shares
but this is less costly when the company is mediocre; hence such ventures will always mimic.
If therefore no separating equilibrium exists then the VC of a good firm will not convert because doing so is not a credible
signal and also results in a lower stake. The proposition clearly depends on the existence of only one available exit strategy and on
the investors' post-investment knowledge (with probability rB) of the venture's true value.
Finally, observe that it is difficult to satisfy the entrepreneur's incentive compatibility condition. The entrepreneur's IC (3) can
be satisfied provided there is a difference in IPO exit probability, between the good and mediocre ventures (i.e. when μ N γ). But
in this case since there is only one VC exit route there is no way to satisfy the entrepreneur's IC and thus no way to motivate her to
exert effort.
5.2. Exit choice without conversion
Consider now the converse situation, where the VC's share f has no conversion rights but he can choose whether to exit via IPO
or TS. In this case, a VC's method of exit can serve to signal the type of his venture to the market.
I analyze whether a separating equilibrium is possible in which the good firm's VC exits through an IPO and the mediocre
firm's VC through a TS. If so, then investors would believe that a firm whose VC exited via IPO (resp. TS) is good (resp. mediocre).
The incentive compatibility conditions for such an equilibrium are:
H
H
f ð1−SH ÞV ð1 þ xÞ þ Z≥f ð1−SL ÞV ð1 þ xÞ
h
i
L
L
H
f ð1−SL ÞV ð1 þ xÞ≥f ð1−SH Þð1 þ xÞ rV þ ð1−r ÞV
ð9Þ
ð10Þ
Inequality (9) is the IC condition for the good venture's VC; it simply states that his overall payoff when exiting through an IPO
is greater than when exiting through a TS. When the good venture's VC exits via IPO he also secures the additional private benefit
Z—the reputation effect described previously. Note that the VC forgoes that benefit Z if he exits via TS.
Similarly, (10) is the IC condition for the mediocre venture's VC. It states that the VC of a mediocre venture does better by
exiting through a TS than through an IPO. A TS exit signifies that the venture is mediocre, so investors pay a price SL for their stake.
In contrast, in case of an IPO investors pay a fair price only with a probability r b 1.
The IC condition of a good venture's VC (inequality (9) will always be satisfied but that of the mediocre venture's VC
(inequality (10) will never be satisfied. Since VH N VL, it follows that, for all values of r b 1, the sum [rVL + (1 − r)VH] will exceed
VL. And since (1 − SH) N (1 − SL) by definition, it follows that the right-hand side of Eq. (10) will always exceed the left-hand
side; and hence the IC condition of mediocre venture's VC can never be satisfied. Consequently, when VCs hold common equity
stakes there is no separating equilibrium whereby VCs of good ventures exit through IPOs while those of mediocre ventures exit
through trade sales. It can likewise be shown that there exists no separating equilibrium in which VC's hold common equity and
the good exit via TS while the mediocre exit via IPO. This result is formalized in our next proposition.
Proposition 2. If VCs hold only common equity, then there is no separating equilibrium in which a VC can signal his venture's type via
choice of exit strategy (either IPO or TS).
Proof. Refer to preceding discussion.
This proposition holds because, for the mediocre venture the VC's IC cannot be satisfied. Unlike the case of conversion
without exit choice, where conversion imposes a cost, in this case the mediocre venture's VC is always better off mimicking the
good venture's VC because that is a no cost strategy. Hence a separating equilibrium cannot exist without conversion and so the
Please cite this article as: Arcot, S., Participating convertible preferred stock in venture capital exits, Journal of Business Venturing
(2013), http://dx.doi.org/10.1016/j.jbusvent.2013.06.001
8
S. Arcot / Journal of Business Venturing xxx (2013) xxx–xxx
VCs of both venture types will choose the same exit strategic; this has a negative effect on the entrepreneur's incentives to exert
effort.
5.3. Exit choice with conversion
Here I explore the possibility of a separating equilibrium when (a) the VC can exit via IPO or TS and (b) the VC's stake is
convertible. Once again I look for a separating equilibrium in which the good venture's VC converts (and exits via IPO) while the
mediocre venture's VC does not convert (and exits via TS). In this case potential investors would believe that the venture is good
only if the VC has converted his stake and exited through an IPO; otherwise the venture is believed to be mediocre. Recall that,
with an IPO exit, investors do not discover the venture's true value even after investment; they observe the true value only with
probability r.
The incentive compatibility conditions for such an equilibrium are as follows:
h
i
H
H
qð1−SH ÞV ð1 þ xÞ þ Z≥f ð1−SL ÞV ð1 þ xÞ
ð11Þ
h
i
L
L
H
f ð1−SL ÞV ð1 þ xÞ≥qð1−SH Þ rV þ ð1−r ÞV ð1 þ xÞ
ð12Þ
For a separating equilibrium to exist the investors must believe that the converting VC's venture is good and will naturally
accept a stake SH in that venture. Condition (11) simply states that the VC of a good venture does better by converting (and thus
accepting a stake q) and paying a price SH for the investment than by not converting (i.e. retaining the stake f), having his venture
be mistaken for a mediocre one and exiting through a TS. Recall that the VC derives private benefit Z when a good firm holds a
successful IPO for a good firm. It should be clear from Eq. (11) that the conversion signal will be credible only if q b f.
Condition (12) states that it is not worthwhile for the mediocre to mimic the good by exiting through an IPO. The VC of a
mediocre venture is better-off not converting (and then exiting through a TS). The reason is that conversion does not guarantee
high valuation of the VC's remaining stake in a mediocre firm. Investors in an IPO realize the venture's true value with probability
r, so it is only with probability (1-r) that the mediocre venture is mistaken for the good. These IC conditions yield the following
proposition
Proposition 3. There exists a fully separating equilibrium—in which the VC of the good venture converts his stake into common equity
and then exits through an IPO—provided that
q∈
f ð1−SL ÞV H ð1 þ xÞ−Z
f ð1−SL ÞV L
;
ð1−SH ÞV H ð1 þ xÞ
ð1−SH Þ rV L þ ð1−rÞV H
Z≥Z min ¼ f
VH
VL
h
L
V ð1 þ xÞ−I
i
ΔV ð1−r Þ
rV þ ð1−rÞV H
iÞ
L
iiÞ
The VC of the mediocre venture does not convert and exits through a trade sale.
Proof. See Appendix A
I define the range in which a fully separating equilibrium exists as q∈ qFS ; qFS . The upper bound of the range i.e. qFS , defines
the threshold above which the mediocre venture's VC no longer finds it worthwhile to exit through a TS. Similarly, below the
lower bound qFS the payoff for a good venture's VC is greater via TS via IPO. Thus it is only when q lies within the range that there
can exist a fully separating equilibria in which the good firm's VC converts his stock (and then exits through an IPO) while a
mediocre firm's VC does not convert (and then exits though a TS). Note that condition (i) of the proposition implies that q b f
hence conversion must be costly (see Appendix A for proof).
The other necessary condition for a separating equilibria to exist is that the private benefits Z can be no less than Zmin as in
condition (ii) where ΔV = (VH − VL). Separation is not possible with Z = 0 because the combination of converting and choosing
an IPO exit imposes a costs on both types of venture. However, the cost is higher for the good ventures than for the mediocre one.
Hence a separating equilibrium can occur only if good firms are more willing than mediocre firms to pay (cf: Allen and Faulhaber,
1989). Yet that can happen only when Z is sufficiently large. The VC of the good firm thus accepts (through conversion) a reduced
stake, and then exits though an IPO because of the reputation (private benefit Z) he gains from bringing a good venture to the
market. This is crucial because—in light of our assumption that trade sales are informationally more efficient than IPOs—the
venture is valued less accurately in the later.
Could there be a separating equilibrium with q = f? Without conversion, the VC of the mediocre firm will be tempted by an IPO
because it would make the venture's true value less detectable than would sticking to his postulated equilibrium strategy of exiting
through a trade sale. Forcing conversion to a lower stake prior to an IPO is one way to deter the mediocre from mimicking the good.
Please cite this article as: Arcot, S., Participating convertible preferred stock in venture capital exits, Journal of Business Venturing
(2013), http://dx.doi.org/10.1016/j.jbusvent.2013.06.001
S. Arcot / Journal of Business Venturing xxx (2013) xxx–xxx
9
The results reported here confirm that exiting through an IPO increases the entrepreneur's incentives. Gompers and Lerner (1999)
point out, IPOs are the most successful of all venture exits and fetch the highest returns to the venture investors. Moreover, most
venture agreements require that—prior to an IPO —convertible stakes be automatically converted into common equity.
5.4. Semi-separating equilibria
In this Section I explore the possibility that a good venture's VC exits via IPO with probability μ and exits via a TS with probability
1 − μ that is a semi-separating equilibria. The conditions for the existence of such a hybrid equilibrium are as follows:
h
i
H
H
qð1−SH ÞV ð1 þ xÞ þ Z≥f 1−S^ V ð1 þ xÞ
h
i
L
L
H
f 1−S^ V ð1 þ xÞ≥qð1−SH Þ rV þ ð1−r ÞV ð1 þ xÞ
ð13Þ
ð14Þ
These conditions are similar to those for the fully separating equilibria; the only difference is that in this case a TS no longer
allows investors to conclude that the venture is mediocre. The investors know that a venture holding an IPO is good, but a venture
raising funds through a TS could be either good or mediocre. Investors update their prior probabilities of good and mediocre i.e. α
and (1 − α) with the additional information about the probability μ of a good venture undertaking an IPO. After Bayesian
updating the investors estimate a revised share S^ for ventures that seek funding via trade sales. This share is given by the
following relationship:
S^
ð1−μ Þα
ð1−α Þ
H
L
V ð1 þ xÞ þ
V ð1 þ xÞ ¼ I
ð1−μ Þα þ ð1−α Þ
ð1−μ Þα þ ð1−α Þ
ð15Þ
Proposition 4. There exists a semi-separating equilibrium in which the VCs of the good firms randomize between converting their PCP
stakes (and exiting through an IPO) and not converting (and exiting through a Trade Sale) —provided
(
q∈
f 1−S^ V H ð1 þ xÞ−Z
ð1−SH ÞV H ð1 þ xÞ
Z≥Z min ¼ f
VH
VL
h
L
;
)
f 1−S^ V L
L
ð1−SH Þ rV þ ð1−r ÞV H
V ð1 þ xÞ−I
i
ΔV ð1−r Þ
rV L þ ð1−rÞV H
1Þ
2Þ
The VCs of the mediocre ventures do not a trade convert and exit only through sale.
Proof. See Appendix A
Much as with the condition
on q for
a fully separating equilibria, we may define the range of q over which a semi-separating
equilibria is supported as q∈ qSS ; qSS . Note that the value of Zmin that supports an equilibrium is the same for both the fully
separating and the semi-separating equilibria. Zmin is the minimum reputation gain required for the good venture's VC to exit through
an IPO. The two types of equilibria differ only in terms of the probability μ of an IPO exit by a good venture. The VCs payoffs from the
IPO exit is the same for both fully separating and semi-separating equilibria, which is why the value of Zmin is the same for both.
If a good venture's VC does not exit through an IPO then his payoff is higher in a semi-separating equilibria than in a fully
separating equilibrium. This difference is reflected in the upper bound of q which is lower in the semi-separating equilibriumqSS b qFS .
Comparing the range of q for a fully separating and semi-separating equilibria reveals that qFS N qSS NqFS N qSS . In sum; values
of q that are low in the range support fully separating equilibria, medium values support both fully separating and semiseparating equilibria, and high values support only semi-separating equilibria.
A natural question is whether the expressions so derived are feasible—in other words do they lead to reasonable values of D, f and
q? To illustrate the answer, a numerical example will prove useful. Assume the following parameters, I = 100, VH = 300, VL =
150, x = 25% and r = 60%. If we assume that Z = Zmin = 50f then using the respective expressions from Proposition 3 yields qFS =
204:56f
0.4545f. Combining qFS and Eq. (1) we have D = ð1−0:4545f
. Thus qFS = 4.55% and D = 21.43 when f = 10% and so on.5 This shows
Þ
that there exist feasible sets of values for the relevant parameters of our model.
More formally using Eq. (1) and the expressions for qFS , we can easily derive an expression for D in terms of f and show that
indeed D N 0 and q N 0 for values of f N 0.
Finally, which separating equilibrium does the VC prefer for medium values of q? Given VH, the probability of exiting through an
IPO is lower in the semi-separating than in the fully separating case. So in terms of incentivizing the entrepreneur, it is optimal—all
else being equal—when VCs commit to exiting through an IPO.
5
In the above example the value of D varies between 21.43 and 132.35 for values of f between 10% and 50% respectively.
Please cite this article as: Arcot, S., Participating convertible preferred stock in venture capital exits, Journal of Business Venturing
(2013), http://dx.doi.org/10.1016/j.jbusvent.2013.06.001
10
S. Arcot / Journal of Business Venturing xxx (2013) xxx–xxx
5.5. Out of equilibrium behavior
In this Section I investigate whether the hypothesized actions taken by the VCs of the good or mediocre ventures dominate the
other actions open to them. To perform this analysis I list below the payoffs to the good and the mediocre VCs under all possible
scenarios i.e. conversion, non-conversion each under the two possible exit routes—IPO and TS.
5.5.1. Good venture
I list below the payoffs to the VC of the good venture taking into account both conversion as well as exit choices.
Conversion
Non-conversion
H
f(1 − SL)(1 + x)[rVH + (1 − r)VL]
f(1 − SL)VH(1 + x)
[q(1 − SH)V (1 + x)] + Z
q(1 − SH)VH(1 + x)
IPO
TS
From the above table it is clear that in case of conversion the VC of the good venture prefers exiting via IPO compared to TS.
Similarly in case of non-conversion he prefers TS to IPO, since if r b 1, then rVH + (1 − r)VL is less than VH(1 + x). In the analysis
of the separating equilibrium (Section 5.3) we have derived parameter values such that the VC prefers conversion and exit via IPO
as compared to non-conversion and exit via TS.
5.5.2. Mediocre venture
The payoffs to the VC of the mediocre venture are:
Conversion
Non-conversion
L
H
q(1 − SH)(1 + x)[rV + (1 − r)V ]
q(1 − SH)VL(1 + x)
IPO
TS
f(1 − SL)VL(1 + x)
f(1 − SL)VL(1 + x)
The VC of the mediocre venture in case of conversion prefers IPO to TS and in case of non-conversion he is indifferent. In proposition 3
we have derived parameter values such that the VC prefers non-conversion and exit via TS to conversion and exit via IPO.
The analysis in this section thus allows us to conclude that for both good and mediocre ventures, the VC's equilibrium choice
(of exit via IPO and TS respectively) is guaranteed not only by their equilibrium behavior but also by some out-of-equilibrium
behavior. Both satisfy standard refinements—including the intuitive criterion of Cho and Kreps (1987), which ensures that the
equilibrium choices dominate the other available options.
6. Discussion
The preceding sections clearly demonstrate that there exist separating equilibria in which the good venture's VC converts and exits
through an IPO while the mediocre venture's VC does not convert and exits through a TS. Given that the entrepreneur's cash flows as well
as private benefits are maximized in an IPO that mode of exit can be used to motivate and reward her efforts, upon which the venture's
success depends. In contrast, the VC of a mediocre venture exits through a trade sale, in which case the entrepreneur receives no private
benefits. The following table summarizes the total payoffs to the VC and entrepreneur in case of a separating equilibrium.
Venture
Payoff
Exit
Value
VC
Entrepreneur
IPO
TS
VH(1 + x)
VL(1 + x)
q(1 − SH)VH(1 + x) + Z
f(1 − SL)VL(1 + x)
(1 − q)(1 − SH)VH(1 + x) + B
(1 − f)(1 − SL)VL(1 + x)
The entrepreneur's payoffs are higher when exit is through an IPO. She is rewarded with higher cash flows as well as private
benefits of control, B. In contrast, when exit is through a trade sale her share is lower, since f N q it follows that (1 − f) b (1 − q)
and moreover she does not get any private benefits of control.
From the VCs viewpoint, even though his share of cash flows realized in an IPO is lower he additionally gets a private benefit Z.
In fact the private benefit should compensate the VC for his loss of cash flow rights upon conversion. The higher is the value of
these private benefits, the lower is the (post-conversion) value q that the VC is willing to accept.
The entrepreneur will always prefer exit via IPO and conversion, since this choice maximizes her cash flows as well as private
benefits. On the other hand, the VCs payoffs are maximized in a trade sale, so his preference will depend crucially on the level of
private benefits (Z), he gets in a successful IPO. To signal the venture's value the VC forgoes his higher share (f) in case of exit via
IPO, which has the effect of incentivizing the entrepreneur. Note that since the VC takes this action only when firm value is high, it
has a lower impact on the VC's overall payoff.
The ex ante optimal values of f and q can be determined based on the expected payoff to the VC, which should be equal to K,
the initial funding requirement of the project.
Please cite this article as: Arcot, S., Participating convertible preferred stock in venture capital exits, Journal of Business Venturing
(2013), http://dx.doi.org/10.1016/j.jbusvent.2013.06.001
S. Arcot / Journal of Business Venturing xxx (2013) xxx–xxx
11
6.1. Conversion
PCP holders face automatic conversion in a qualified public offering (QPO) which is an initial public offering (IPO) of securities
valued at or above a total amount specified in the financing agreement. This amount is usually specified to be sufficiently large
enough to guarantee that the IPO shares will trade on a major exchange. Kaplan and Strömberg (2003), document that this
automatic conversion rule typically applies only if the company completes an IPO at a share price that exceeds that of the previous
financing round by a factor (on average) of 3. This 3:1 ratio is significantly higher if the share price at prior investment rounds is
also considered. Thus the VC is not prepared to give up both control and participation/preferred rights unless he is assured of a
high exit value.
Most VC financing agreements explicitly indicate what conditions lead to conversion of the VC stake, but this paper models the
decision to convert as a choice made by the VC at the time of his exit. Modeling that choice as an ex ante decision would not alter
the implications of the signaling equilibria described here, since in either case the VC forgoes his participation rights in a public
offering but not in a trade sale. The VC's agreeing ex ante to give up those rights is essentially a commitment device to avoid ex
post conflicts.
6.2. Alternative explanations and securities
Our model focuses on one reason for the use of participating convertible preferred stock in venture capital contracts and their
conversion prior to a public offering. Of course, there may be other reasons for the presence of PCP stock in this context. Some of
these reasons are discussed in this subsection.
A major reason for use of PCP stock is the protection it affords VCs from unscrupulous entrepreneurs. This motivation is
evidenced in the Hotmail Corporation case study (Mukherjee (1999)). During negotiations with the VC the entrepreneurs were
concerned that the investors were receiving participating preferred stock in their first round of investment. The entrepreneurs felt
that this was unfair for the investors to thus “double-dip”—which means that they recovered their original investment and shared
in the remaining equity pool. The VCs in this case study countered that the arrangement was necessary so that entrepreneurs
would not have an incentive to sell the company (at a low price) early in its life.
Practitioners offer another reason for the forced conversion of PCPs in an IPO. Venture capitalists convert their preferred
holdings to common shares in order to clean up the company's capital structure before it goes public. Otherwise that structure
would be so complicated that potential IPO investors would find it too costly to perform their due diligence. These concerns do
not arise in a trade sale because the investor acquires the entire company; the VC and the entrepreneur are then left to distribute
the proceeds among themselves.
In many cases the preferred shares have a cap; these are known as participating convertible preferred with cap (PCPC) shares.
With such shares, the total of the liquidation preference and participation in excess earnings is capped at a pre-specified level. The
cap is usually stipulated as a multiple of the liquidation preference value. In order to receive a payout in excess of the specified cap
the PCPC shareholder would have to convert that stake into common equity. How does this class of stock affect our model? The
cap plays a role similar to that of a QPO. A VC will not convert unless he is assured of a high return—irrespective of the mode of
exit. When that exit is via TS, however, there is no signaling role for the preferred shares. Even so, the basic point remains that a
high valuation has been achieved, which is probably why the VC is willing to forgo his participation rights.
The analysis presented in this paper brings up one additional question: Can we use a simple convertible preferred (CP) stock
instead of the PCP shares described here—to achieve the same results? It is certainly true that for exit valuations below a certain
threshold, CP and PCP securities would yield similar results. This claim can be illustrated using numbers from my example in the
Introduction. For an exit valuation below $10 million (say $9 m), CP holders would be entitled to their preferred dividend of $5 m
when not converting or to $4.5 m (50% of $9 m) upon conversion. However, this generalization applies only for low exit
valuations. The focus in this paper is on exits via IPOs and trade sales, transactions that almost always feature extremely high
valuations. Moreover, even with low valuations it costs more to signal using PCP than CP stock; this means that the former
security type supports a separating equilibrium over a greater range of q values. The empirical evidence indicates that in about
half of all cases, convertible preferred stock is nonparticipating. One possible occasion for the use of nonparticipating stock is
because of uncertainty about the entrepreneur's effort. Nevertheless, the choice between participating and nonparticipating
convertibly preferred stock raises interesting issues worth exploring.
Almost all preferred securities issued to VCs (including CP, PCP and PCPC) are subject to clauses that require conversion into
common prior to an IPO. However, preferred securities differ from other securities in that converting the latter is a decision that
simply involves comparing payoffs and then choosing the highest one. In contrast, a VC who converts a PCP stake is giving up his
participation rights—which can happen only with PCP stock and is the cost of signaling in our model.
6.3. Optimal security design
In the model I have abstracted from the issue of optimal security design. We have seen that, for a separating equilibrium to
exist it is necessary that q b f. One way to ensure this outcome is through the participating and convertibility features of PCP stock.
However, I have not addressed the question of how best to design the initial stake.
Given our model's two point support (VH and VL) our focus on equity is without loss of generality. A combination of equity
and risky debt can replicate the payoffs from any optimal security. It should therefore be straightforward to extend our
Please cite this article as: Arcot, S., Participating convertible preferred stock in venture capital exits, Journal of Business Venturing
(2013), http://dx.doi.org/10.1016/j.jbusvent.2013.06.001
12
S. Arcot / Journal of Business Venturing xxx (2013) xxx–xxx
argument to the case where a VC holds both debt and equity and could convert (at some cost) one or both securities in the event
of an IPO.
A more general conclusion that can be drawn from my analysis is that the informed party (in our case the VC) should hold a
claim that is sufficiently information sensitive that converting it to a lower stake conveys some information. I posit the
existence of a trade-off between this claim's information sensitivity and how much a VC must forgo in order for such action to
be viewed as a credible signal. In a more inclusive model, the issues of security design and convertibility would be considered
jointly. The key point here is simply that the optimal security should be convertible otherwise it cannot be used to signal
value.
Unlike many other models in the literature, my model does not incorporate the double moral-hazard setting. In that
setting, not only does the entrepreneur typically faces some moral hazard problem (as in this paper), but also the VCs
must be given incentives because their actions, too are non-contractible. Although model details vary, most theoretical
work reaches the conclusion: that convertible preferred equity enables the optimal incentives for both entrepreneur and
VC.6
Finally, recent VC literature has analyzed how behavioral factors influence the VC-entrepreneur relationship. For example Fairchild
(2011), develops a game-theoretic model with both economic and behavioral factors in a double-sided moral hazard setting. The
entrepreneur in his model—when choosing between “ängel” and venture capital financing—weighs her empathy with the angel against
the VC's value adding abilities. We can reasonably conclude that including behavioral factors would affect the contracts and thus the
results discussed in this paper. Such factors thus provide an interesting avenue for extending our work.
7. Empirical implications
The first question that comes to mind is; Under what circumstances are we likely to observe to financing that involves PCP
shares? A general feature of such stock (one that our model does not exploit) is that their participation rights facilitate
overcoming financial constraints and thereby help support higher investments. This means that PCP issues are more likely to be
used in relatively larger projects. More specifically; although the model developed here does not explicitly compare the use of
PCPs and other securities, however I do show that participating stock can be used to signal value when a firm's VC exits through
an IPO. As a result, PCP stock is likely to be used whenever an IPO exit is possible. For this reason, their use is more likely when the
venture's success is uncertain and contingent on entrepreneurial effort.
In what follows I use comparative statics to analyze how changes in different parameter values affect the range of q-values that
support a separating equilibrium. Then the results of the analysis are used to make empirical predictions based on those. I focus
on two parameters Zmin and q that define the equilibrium.
Lemma 5. The minimum private benefit Zmin required to sustain a separating equilibrium (i) increases with (1-r), with the difference
ΔV between the values of the good and mediocre ventures, with the value VH of the good firm, and with the VC's stake q after
conversion and (ii) decreases with value VL of the mediocre venture and with the amount I of investment raised.
h
i
Proof. Rearranging the definition of Zmin in Proposition 3(ii) yields Z min ¼ Vq V H ð1 þ xÞ−I ð1−r ÞΔV. From this equality, the
comparative statics follow.
L
The interpretation offered by this paper is that a VC's private benefits Z are reputation effects that help the VC establish himself
among investors. The range of q that supports a separating equilibrium is wider if the distribution of Z is higher in the sense of
first-order stochastic dominance. In the dominant distribution it is more likely that Z ≥ Zmin, which makes it easier to give
entrepreneurs the incentive of a future IPO. In practice, Z is likely to be higher for younger VCs (who have yet to establish their
reputations) and lower for older, more established VCs. Hence the model predicts that we will observe more exits through IPO
exits by younger than by older venture capitalists. This prediction is confirmed empirically by Gompers (1996), who observes
“grandstanding” by younger VCs. Gompers finds that younger VCs are more likely to exit a venture through an IPO than are older
VCs.
Lemma 6. In a fully separating equilibrium the range of q (or f) supporting the equilibria (i) increases as r increases; (ii) increases as Z
increases; (iii) decreases as VH increases; and (iv) decreases as VL increases.
Proof. See the Appendix A
With increasing probability r that a venture's true value is known, the range supporting a separating equilibrium also
increases. It is safe to assume that r is high for informationally efficient markets (e.g. the US and UK markets) and lower for less
efficient ones (as in, for example, much of Continental Europe). In informationally efficient markets there are many analysts
6
There is an extensive literature on double moral hazard in the venture capital context. Notable contributions include Casamatta (2003) and Repullo and
Suarez (2004), for double-sided moral hazard combined with doubles sided asymmetric information see Houben (2003). Excellent reviews are offered by Da Rin
et al. (2011)) and Fairchild(2011).
Please cite this article as: Arcot, S., Participating convertible preferred stock in venture capital exits, Journal of Business Venturing
(2013), http://dx.doi.org/10.1016/j.jbusvent.2013.06.001
S. Arcot / Journal of Business Venturing xxx (2013) xxx–xxx
13
following stocks and so the probability of correctly valuing a post-IPO stock is relatively high, implying a high r. In fact, the
cross-listing literature provides evidence that this is indeed the case. Baker, Nofsinger and Weaver (2002) find that firms listing
their shares on both the New York Stock Exchange (NYSE) or the London Stock Exchange (LSE) and their home country
experience a significant increase in visibility, as proxied by both analyst coverage and print media attention. Other studies
analyzing cross-country differences in, cost of capital (Hail and Leuz (2006)), and IPO underpricing (Banerjee et al. (2011) reach
similar conclusions about the US and other equity markets. We are thus likely to observe more IPO exits in such markets (and
conversely, fewer IPO exits in less visible markets). Such correlations might account for more IPO exits being observed in the US
than in any other market. Also for extremely low values of r (i.e. r → 0) a separating equilibrium cannot be sustained at all. In that
case the use of convertible securities is redundant. This claim also is borne out by empirical and anecdotal evidence. Note that
there is limited use of such securities outside of the United States, especially in Europe (Schwienbacher (2008)). An increase in r
means that the probability that investors know the venture's true value, has the effect of “loosening” the incentive compatibility
condition of the mediocre venture's VC. In the limit when r = 1 investors are certain of the mediocre venture's true value. In
effect, this certainty increases the range over which a separating equilibrium is supported.
In sum; our analysis suggests that we are more likely to observe the use of participating shares in well-developed capital
markets and for larger projects whose success is both uncertain and contingent on entrepreneurial effort.
8. Conclusion
This paper presents a signaling model of exits by venture capitalists. I argue that participating convertible preferred
securities, a sophisticated contracting device, can be used by a VC to signal the venture's quality at the time of exit. Exit
through an IPO has the advantage of incentivizing entrepreneurs, who are usually rewarded with control after the exit. In
contrast, the entrepreneur normally loses control of her venture when the VC exits through a trade sale. However, the VC
views an IPO as having an informational disadvantage when compared with a TS. Hence a VC may well be reluctant to exit via
IPO.
That mismatch of preferences can be resolved with the aid of PCP stock. The VCs conversion of his PCP stake into common
equity serves to signal the quality of his venture. In this paper I show that there exists a separating equilibrium in which the VC of
a good venture converts his stake (and exits through an IPO) while the VC of a mediocre venture does not convert (and exits
through a TS). This equilibrium rewards the entrepreneur with control if the venture is a good one. I thus explain a commonly
observed phenomenon in venture capital exits; conversion of VC's stake prior to an IPO.
The model's parameters can also be used to explain some commonly observed differences between the United States and
other VC markets. In particular, given that U.S. markets feature less informational asymmetry, our model predicts that the United
States witnesses more exits through IPOs than do other markets. Finally, we have seen that reputation effects lead to IPO exits
being chosen more frequently by younger than by older venture capitalists.
Appendix A
Proof of Proposition 1. The incentive compatibility condition (7) for the good can be simplified and rewritten as follows
h
i
r B V H þ ð1−r B ÞV L
qð1−SH Þ
≥
f ð1−SL Þ
VH
Similarly the IC (8) for the mediocre can be rewritten as
qð1−SH Þ
VL
≤
L
f ð1−SL Þ
r B V þ ð1−r B ÞV H
The above conditions will be satisfied only if
h
i
H
L
r B V þ ð1−r B ÞV
VL
≥
r B V L þ ð1−r B ÞV H
VH
Simplifying the above leads us to the following condition
H
L 2
0≥ V −V
which can never be satisfied for any values of VH and VL,which means that there exists no separating equilibrium.
Please cite this article as: Arcot, S., Participating convertible preferred stock in venture capital exits, Journal of Business Venturing
(2013), http://dx.doi.org/10.1016/j.jbusvent.2013.06.001
14
S. Arcot / Journal of Business Venturing xxx (2013) xxx–xxx
Proof of Proposition 3. The condition (11) for the good can be simplified and rewritten as
q≥
h
i
f ð1−SL ÞV H ð1 þ xÞ −Z
ð16Þ
ð1−SH ÞV H ð1 þ xÞ
Similarly Eq. (12) can be rewritten as
q≤
f ð1−SL ÞV L
ð1−SH Þ rV L þ ð1−r ÞV H
ð17Þ
Thus a fully separating equilibrium exists if q lies within the values shown above. This gives the first condition for the existence
of the fully separating equilibria. We derive below the minimum value of Z, Zmin which ensures that q lies in the range described
by Eqs. (16) and (17). The above conditions imply that
h
i
f ð1−SL ÞV H ð1 þ xÞ −Z
f ð1−SL ÞV L
≥
ð1−SH Þ rV L þ ð1−r ÞV H
ð1−SH ÞV H ð1 þ xÞ
ð18Þ
Rearranging and simplifying Eq. (18) gives us the minimum value of Z, Zmin for which a fully separating equilibria exists.
I now show that f N q in the above equilibrium. Rearranging Eq. (17) I have
f≥
h
i
qð1−SH Þ rV L þ ð1−rÞV H
From the above f N q if
ð1−SH Þ
ð1−SL Þ
ð19Þ
ð1−SL ÞV L
ð1−SH Þ½rV L þð1−r ÞV H ð1−SL ÞV L
½
Þ
N1, which I can rewrite as ðð1−S
1−S Þ H
L
rV L þð1−rÞV H VL
.
is greater than 1 since by definition SL N SH.
Also, ½
rV L þð1−r ÞV H VL
N1, since rVL + (1 − r)VH N VL. Therefore f N q.
Proof of Proposition 4. I can arrive at the range which supports the semi-separating equilibrium using a similar method used above
for the fully separating equilibrium. The condition (13) for the good can be simplified and rewritten as
q≥
h i
f 1−S^ V H ð1 þ xÞ −Z
ð1−SH ÞV H ð1 þ xÞ
ð20Þ
Similarly Eq. (14) can be rewritten as
f 1−S^ V L
q≤
ð1−SH Þ rV L þ ð1−r ÞV H
ð21Þ
Thus a semi-separating equilibrium exists if f lies within the values shown above. This gives the first condition for the existence of the
semi-separating equilibria. Similarly I derive below the minimum value of Z, Zmin which ensures that f lies in the range described by Eqs.
(20) and (21). The above conditions imply that
h i
f 1−S^ V L
f 1−S^ V H ð1 þ xÞ −Z
≥
ð1−SH Þ rV L þ ð1−r ÞV H
ð1−SH ÞV H ð1 þ xÞ
ð22Þ
Rearranging and simplifying Eq. (22) gives us the minimum value of Z, Zmin above which the semi-separating equilibria exists.
Proof of Lemma 7. The comparative statics of the various parameters supporting the fully separating equilibrium is arrived at
by differentiating the upper and lower bounds with the respective parameters. For ease of exposition I calculate the impact on
the range by restating and differentiating the upper and lower bounds w.r.t. f. The results hold if the effects are calculated
w.r.t. q.
Please cite this article as: Arcot, S., Participating convertible preferred stock in venture capital exits, Journal of Business Venturing
(2013), http://dx.doi.org/10.1016/j.jbusvent.2013.06.001
S. Arcot / Journal of Business Venturing xxx (2013) xxx–xxx
15
(i) With respect to r
qð1−S Þ V −V
Differentiating the lower and upper bound of f with respect to r I have ∂ ∂rf ¼ ð1−S½ ÞV which can be further simplified as
∂ f FS
∂r
¼
−qð1−SH ÞV ½V −V
H
L
2
ð1−SL Þ½rV L þð1−rÞV H L
H
FS
L
L
H
L
. Thus ∂ ∂rf b0. Also the upper bound f FS does not depend on r and hence does not change with r. Thus an
FS
increase in r results in the lower bound fFS decreasing which causes the entire range to increase.
(ii) With respect to Z
ÞV ð1þxÞþZ
Only the upper bound f FS depends on Z. On inspecting the value f FS ¼ qð1−S
I can immediately see that an increase in
ð1−S ÞV ð1þxÞ
H
H
L
H
Z increases f FS . The value of fFS does not depend on Z and thus is not affected by it. Therefore an increase in Z results in an
increase in the range of f supporting the equilibria.
(iii) With respect to VH
Differentiating the upper and lower bounds of f with respect to VH I have the following:
i
2h L
3
H
∂ f FS
q 4I rV þ ð1−r ÞV
¼
þ ð1−SH Þð1−r Þ5
H
ð1−SL Þ
∂V H
V ð1 þ xÞ 2
qI−Z
which is b 0. Thus an increase in VH leads to an increase in fFS and
which is clearly N 0. Thus ∂∂Vf N0. Similarly ∂f∂V ¼
ð1−S Þ½V ð1þxÞ
a decrease in f FS which has the effect of decreasing the range supporting the equilibria.
(iv) With respect to VL
Differentiating the upper and lower bounds of f with respect to VL I have the following:
FS
H
FS
H
L
H 2
"
#
(
)
∂ f FS
I
VH
ð1−r Þ V H
¼
q
ð
1−S
Þ
r
þ
ð
1−r
Þ
−
L 2
H
ð1−SL Þ
∂V L
VL
ð1−SL Þ2 V L ð1 þ xÞ 2
V
The above expression is positive if
I
2
ð1−SL Þ2 V L ð1 þ xÞ
"
r þ ð1−r Þ
#
H
H
V
ð1−r Þ V
:
≥ 2
L
V
V L ð1−SL Þ
Simplifying it can be shown that it is not true, thus
∂ f FS
∂V L
b0.Similarly,
∂f FS
qð1−SH ÞV H ð1 þ xÞ þ Z
I
1
¼−
L
∂V
V H ð1 þ xÞ
ð1−SL Þ2 V L ð1 þ xÞ 2
which is clearly b 0.
Thus an increase in VL causes both the lower bound and upper bound to decrease resulting in a decreased range of f supporting the
separating equilibrium.
References
Aghion, P., P. Bolton and J. Tirole, 2000, Exit options in corporate finance: liquidity versus incentives, mimeo, University of Toulouse.
Allen, F., Faulhaber, G.R., 1989. Signalling by underpricing in the IPO market. Journal of Financial Economics 23, 303–323.
Amit, R., Brander, J., Zott, C., 1998. Why do venture capital firms exist? Theory and Canadian evidence. Journal of Business Venturing 13, 441–466.
Baker, K.H., Nofsinger, J.R., Weaver, D., 2002. International cross-listing and visibility. The Journal of Financial and Quantitative Analysis 37 (3), 495–521.
Bascha, A., Walz, U., 2001. Convertible securities and optimal exit decisions in venture capital finance. Journal of Corporate Finance 7 (3), 285–306.
Banerjee, S., Dai, L., Shrestha, K., 2011. Cross-country IPOs: what explains differences in underpricing? Journal of Corporate Finance 17, 1289–1305.
Bayar, O., Chemmanur, T., 2011. IPOs versus acquisitions and the valuation premium puzzle: a theory of exit choice by entrepreneurs and venture capitalists.
Journal of Financial and Quantitative Analysis 46 (6), 1755–1793.
Berglöf, E., 1994. A control theory of venture capital finance. Journal of Law, Economics and Organization 10 (2), 247–267.
Bienz, C., and T. Leite, 2008, A Pecking-Order of Venture Capital Exits, mimeo, Norwegian School of Economics and Business Administration (NHH).
Bienz, C., Walz, U., 2007. Evolution of decision and control rights in venture capital contracts: an empirical analysis. RICAFE Working Paper No. 040. London School
of Economics.
Black, B.S., Gilson, R.J., 1998. Venture capital and the structure of capital markets: banks versus stock markets. Journal of Financial Economics 47 (3), 243–277.
Casamatta, C., 2003. Financing and advising: optimal financial contracts with venture capitalists. Journal of Finance 58 (5), 2059–2086.
Chemmanur, T.J., Fulghieri, P., 1999. A theory of the going-public decision. Review of Financial Studies 12 (2), 249–279.
Cho, I., Kreps, D.M., 1987. Signalling games and stable equilibria. Quarterly Journal of Economics 102 (2), 179–222.
Cumming, D.J., MacIntosh, J.G., 2003. A cross-country comparison of full and partial venture capital exits. Journal of Banking & Finance 27, 511–548.
Da Rin, M., Hellmann, T., Puri, M., 2011. A survey of venture capital research. TILEC Discussion Paper No. 2011-044. Tilburg Law and Economics Center, Tilburg
University.
Fairchild, R., 2011. An entrepreneur's choice of venture capitalist or angel-financing: a behavioral game-theoretic approach. Journal of Business Venturing 26,
359–374.
Faure-Grimaud, A., Gromb, D., 2004. Public trading and private incentives. Review of Financial Studies 17 (4), 985–1014.
Gompers, P.A., 1996. Grandstanding in the venture capital industry. Journal of Financial Economics 42, 133–156.
Gompers, P.A., Lerner, J., 1999. The Venture Capital Cycle, Cambridge. MIT Press, Mass.
Please cite this article as: Arcot, S., Participating convertible preferred stock in venture capital exits, Journal of Business Venturing
(2013), http://dx.doi.org/10.1016/j.jbusvent.2013.06.001
16
S. Arcot / Journal of Business Venturing xxx (2013) xxx–xxx
Hail, L., Leuz, C., 2006. International differences in the cost of equity capital: do legal institutions and securities regulation matter? Journal of Accounting Research
44 (3), 485–531.
Hellmann, T., 2006. IPOs acquisitions and the use of convertible securities in venture capital. Journal of Financial Economics 81 (3), 649–679.
Houben, E., 2003. Venture capital, double-sided adverse selection and double-sided moral hazard. Working Paper No.556. University of Kiel.
Kaplan, S., Strömberg, P., 2003. Financial contracting meets the real world: an empirical study of venture capital contracts. Review of Economic Studies 70 (2),
281–315.
Kaplan, S., Sensoy, B.A., Strömberg, P., 2009. Should investors bet on the jockey or the horse? Evidence from the evolution of firms from early business plans to
public company. Journal of Finance 64 (1), 75–115.
Megginson, W.L., Weiss, K.A., 1991. Venture capitalist certification in initial public offerings. Journal of Finance 46 (3), 879–903.
Metrick, A., Yasuda, A., 2011. Venture Capital & The Finance of Innovation, Second edition. John Wiley & Sons, Inc.
Mukherjee, P., 1999. Hotmail Corporation. Graduate School of Business, Stanford University, Case Study.
Poulsen, A.B., Stegemoller, M., 2008. Moving from private to public ownership: selling out to public firms versus initial public offerings. Financial Management 37,
81–101.
Repullo, R., Suarez, J., 2004. Venture capital finance: a security design approach. Review of Finance 8 (1), 75–108.
Schwienbacher, A., 2008. Venture capital investment practices in Europe and the United States. Financial Markets and Portfolio Management 22 (3), 195–217.
Tykvová, T., 2003. The decision of venture capitalists on timing and extent of IPOs. Discussion Paper 03-12. Centre for European Economic Research (ZEW).
Zingales, L., 1995. Insider ownership and the decision to go public. Review of Economic Studies 62 (3), 425–448.
Please cite this article as: Arcot, S., Participating convertible preferred stock in venture capital exits, Journal of Business Venturing
(2013), http://dx.doi.org/10.1016/j.jbusvent.2013.06.001